The overall hypothesis of this grant is that the local control of lung cancer can be significantly improved, at fixed or reduced morbidity levels, by advances in three- dimensional imaging, treatment planning, and delivery which overcome dose delivery errors due to breathing motion. A critical component of this goal is the quantitative understanding of how the lung and lung tumor tissues move as the patient breathes. Treatment planners have to increase the size of the radiation portal to account for tumor motion, engage the radiation beam only when the tumor is beneath the portal, or track the portal with the tumor motion. Each of these strategies requires that the tumor and other lung tissue positions be accurately known. Our group has developed a novel mathematical lung tissue and lung tumor motion (trajectory) model that relates the tissue positions to the breathing depth and rate of breathing. Specific aim 1 will develop and validate an image deformation algorithm, coupled with computed tomography imaging and reconstruction methods, to provide optimal input data for the breathing motion model. Specific aim 2 will show that our motion model does not change appreciably for patients that do not have their lungs irradiated. Specific aim 3 will show that, for lung cancer radiation therapy patients, the lung motion model will change during therapy. We will evaluate the changes and hypothesize that the change can be modeled during the course of therapy, allowing the treatment planner to plan for future changes (adaptation) rather than assume that no change takes place. In addition to radiation therapy, the breathing motion model has potential applications in nuclear medicine imaging and lung physiology research. PUBLIC HEALTH RELEVANCE: We are developing a mathematical description of human breathing motion for use in radiation treatments of lung cancer. The description uses imaging data acquired using sophisticated computed tomography (CT) scans that are acquired while the patient's breathing is monitored. The CT acquisition and analysis methods will be developed to maximize the model performance and the performance will be measured in lung-cancer and non-lung cancer radiation therapy patients. If the motion model is found to be accurate, it may have applications outside of radiation therapy.